RFR: 8272138: ZGC: Adopt release ordering for self-healing [v3]

[Xiaowei Lu]

On Tue, 10 Aug 2021 14:18:36 GMT, Per Liden <pliden at openjdk.org> wrote:

>> src/hotspot/share/gc/z/zRelocate.cpp line 77:
>> 
>>> 75:   ZUtils::object_copy_disjoint(from_addr, to_addr, size);
>>> 76: 
>>> 77:   OrderAccess::release();
>> 
>> Could you please replace the calls to `OrderAccess::release()` with a single call here instead:
>> 
>> ```--- a/src/hotspot/share/gc/z/zForwarding.inline.hpp
>> +++ b/src/hotspot/share/gc/z/zForwarding.inline.hpp
>> @@ -136,6 +136,8 @@ inline uintptr_t ZForwarding::insert(uintptr_t from_index, uintptr_t to_offset,
>>    const ZForwardingEntry new_entry(from_index, to_offset);
>>    const ZForwardingEntry old_entry; // Empty
>>  
>> +  OrderAccess::release();
>> +
>>    for (;;) {
>>      const ZForwardingEntry prev_entry = Atomic::cmpxchg(entries() + *cursor, old_entry, new_entry, memory_order_relaxed);
>>      if (!prev_entry.populated()) {
>> 
>> 
>> While keeping the calls close to the copy code has some value, I think I'd prefer to have it in a single place, and close to the corresponding acquire in `at()`.
>
> I updated my comment above, since it had a typo (used `memory_order_release` instead of `memory_order_relaxed`).

OK, it's better to add the fence in one single place.

-------------

PR: https://git.openjdk.java.net/jdk/pull/5046